Feeder-level fault detection and classification with multiple sensors: a smart grid scenario

Abstract

The smart grid initiative requires self-healing distribution systems with more accurate fault detection and classification techniques. A multi-sensor feeder-level fault detection and classification algorithm is presented in this work, based on the techniques of the support vector machine and the principal components. An IEEE 34-bus feeder model with dynamic loading conditions is used to evaluate the developed algorithm. Noise in the three-phase current measurements is applied. The numerical analysis indicates that high accuracies in fault detection and classification are achieved for the proposed algorithm.